The completed life-cycle Union Army data set is based on 35,570 men out of a randomly drawn sample of 39,616 males who were mustered into the Union Army during 1861-1865. Socioeconomic and biomedical histories of the recruits from childhood to death have been created by linking together information from different sources. The specific aims of this program project are: 1) to investigate the impact of socioeconomic and biomedical insults duringdevelopmental, middle-life, and older ages on the onset of specific chronic diseases at middle and late ages, on the capacity to work during these ages, on the demand for retirement, and on waiting time to death from specific causes; 2) to chart the way in which these life-cycle interactions have changed with successive cohorts that reached age 65 during the twentieth century; 3) to chart the way in which these life-cycle interactions have changed for different races and different socioeconomic groups; 4) to study the impact of familial factors (environmental and genetic elements taken together) on mortality, health in later life, and exceptional longevity; 5) to estimate what types of public health interventionshave been most effective in lowering mortality rates and improving health; 6) to expand the range of biomedical and socioeconomic factors that can be considered in the life-cycle Union Army data set by linking it to the 1880 census, which provides health information, and by linking it to data on the epidemiological characteristics of cities and of wards in the late 1800s and early 1900s and 7) to create three new life-cycle samples, one for black Union Army veterans, one for men rejected for service from the Union Army, and one for Union Army veterans found in a large, previously collected data set gathered from published U.S. family histories. The first two data sets will permit us to examine populations not represented in the life-cycle Union Army data set, the second sample will permit us to consider more fully racial differences in the aging process, and the third sample will allow us to control for familial characteristics